Synthesis and Development of Deep Eutectic Solvents: Characterization Using Quantitative Structure-Property Relationship, and Application as New Solvents in Diesel Purification

Abstract

The main objective of this thesis is to develop novel models to predict the physio-chemical properties of DESs using the quantitative composition-property relationship (QSPR). The models were developed using the two methods, multiple linear regression (MLR) and artificial neural network (ANN). From the literature, a data set of more than 100 DES and more than 2500 experiment points measuring the physicochemical properties of these DESs, including density, viscosity, electrical conductivity, and pH were collected. The results showed that the proposed models are able to predict the properties of DESs with very high accuracy and can be used for their determination in the absence of experimental measurements, hence reducing the cost, and time for an optimal process design. In addition to this, DESs were investigated for their extraction capacity of thiophene, pyridine, pyrrole, and toluene from n-decane via liquid−liquid extraction (LLE). First, the selected DESs were characterized by measuring their density, dynamic viscosity, and water content. Then, the solubility of each fuel impurity in the DESs was measured. Moreover, the liquid−liquid equilibrium (LLE) data of the pseudo-ternary systems {n-decane (1) + thiophene/pyridine/pyrrole/toluene (2) + DESs (3)} were determined at 298.15 K and 1.01 bar. The solute distribution ratios, selectivities, and the extraction efficiencies of each impurity at a 1:1 solvent to-feed mass ratio were calculated from the experimental LLE data and compared to a benchmark solvent, other ionic liquids (Ils), and DESs reported in the literature. Based on the obtained results, it was concluded that DESs could be considered as effective solvents for the extraction of fuels impurities and can therefore be used in an industrial zone.